GPT-5.2 Support in Warp & New Terminal-Bench Score

Warp has fully integrated OpenAI's GPT-5.2, offering enhanced coding performance to its users after rigorous benchmarking and testing. GPT-5.2 excels in planning, speed, reliability, and completing complex tasks, allowing it to propose coherent multi-step solutions with minimal guidance and verify changes more reliably, reducing dead-ends. The model demonstrates faster execution and better user experience, employing parallel tool usage for efficient searches and edits, resulting in a smoother interaction compared to earlier versions. Achieving a notable score of 61.14% on Terminal Bench, GPT-5.2, in conjunction with Warp's agent platform, marks Warp's strongest performance, highlighting the depth of integration and collaboration with OpenAI to optimize the model for terminal use. Users are encouraged to experience GPT-5.2 in Warp and provide feedback.